Mark Craddock of CrossPlanes asked a good question about two issues with Cypher System that sometimes bug people:
"Do your players mind sharing their hit points with resources in the Cypher System, and have you ever thought of switching from a D20 to a D10 and not multiplying by 3?"
I responded on G+ but since that's highly ephemeral and soon to stop existing a good response should be here, too.
So the D20 vs. D10 idea could work but I like the D20 because it provides a bit more nuance if needed. Some minor rules situations with the nuance include the fact that the D20 allows for a 17-20 escalating level of critical success, and a 1 is a forced GM Intrusion but only has a 5% chance of happening. On a D10 that would at minimum require either require a 1 to be a 10% chance of failure or prompting a second D10 roll to confirm it...and you'd have to revamp how crit ranges work since under this system you'd have crits on a 9-10 instead of 17-20. Not out of the question, but for me these are pretty important parts of the D20 range in play.
I think I'd be perfectly open to the concept of a D10-based value, but I think the intent of the system was to have a 1-30 range of difficulty, but to keep it in discreet "3 point increments" using the level mechanic so that there isn't a temptation for the game to get overly granular. You get a decent numerical spread, but lock down how much you can manipulate that range. That said....I'd take a D10 model over a version that let people spend single point increments to reduce task ranges on a 1-30 scale.
On the idea of health and stat pools being one and the same, my group is not bothered by it and accused me of metagame think when I brought it up (because it was something I had considered). My thought was: there will be on occasion attacks you know deal less damage than the cost to reduce risk, so it's sometimes more sensible to avoid spending points when the cost to spend is lower than the incoming damage risk. My players pointed out to me that it really isn't so evident on their end that these cases may happen, and that the system as written, with the level mechanic for foes, allows them to gauge risk from a more organic level, while insuring they don't need to worry about spending from the pool unless it really counts. So for example, when they know the foe is level 2, they rarely sink points in to it and in fact may find their raw assets and edge let them gain significant or even automatic advantage in those situations. I in turn tend not to throw level 1-3 encounters at the party to waste their time if these fights are going to be trivial.
A more interesting consideration on the resource pool as health is that it also is fatigue. The resource pool is your hard cap on what you can accomplish in a session, barring a change to do recoveries, and as a result this is the first RPG I have ever played in which players are by default tracking fatigue whether they realize it or not. I like that feature a lot.
I also like the fact that a player can try to play it safe if they want and have a character who does not engage in much risk....they roll against the flat numbers and avoid spending the points. It's basically a playstyle concession, and suggests that rather than a character's traits being "always on" like in other RPGs, in Cypher you don't get things done (with a measure of success) unless you put the effort in to it, and as a result, you can be a competent person who doesn't "try hard" essentially.
If there is one key issue with health as resource pools, it is that warriors don't make out as well in this regard. If you are a warrior, you might have might-based effects which cost points, but you also need those points in order to, you know, live. The good news is, most warrior talents seem to be low cost and you can still use Speed as your attack skill. The bad news is, most physical damage defaults to Might damage first.
Cypher could probably benefit from a hybrid system where it has a health pool, and when that pool hits certain points it costs extra effort for the resource pools, but that does lead to extra stat tracking. The current system works well in that it doesn't require anything extra outside of the core resource pool economy to track.